Grow your database from gigabytes to terabytes without changing a line of application code. Transparent sharding handles the scaling.
Start Free Trial
Postgres is the workhorse database for modern applications. But at scale, it breaks in the same predictable ways:
Most teams choose between two bad options: build custom sharding (months of work, months of maintenance) or pay cloud vendors premium rates for distributed SQL.
Neki is Postgres - with transparent sharding built in. Your application sees a single Postgres instance. We handle the rest.
Your code stays unchanged. Neki intercepts queries, routes them to the right shard, and reassembles results. No ORM gymnastics, no custom routing logic.
Add shards as you grow. Rebalance keys across shards in the background while traffic continues. No scheduled maintenance windows.
Joins, subqueries, and aggregates work across shards. The query planner understands your shard key and optimizes accordingly.
Each shard is replicated. Failover happens in seconds, with no manual intervention or query routing changes required.
See shard health, data distribution, and growth trends. Drill down into per-shard query metrics and hot spots.
Every shard rebalance, schema change, and failover is logged. Understand what happened, when, and why.
Choose a column that partitions your data naturally (user_id, organization_id, customer_id). Neki distributes data across shards by this key.
Run Neki as a managed service or self-hosted. Point your application connection string at Neki instead of Postgres. That's it.
As data grows, Neki splits shards automatically. Queries work the same way. Your application never knows sharding happened.
Neki acts as a Postgres-compatible proxy. When a query arrives:
| Feature | Custom Sharding | CloudSQL/RDS Read Replicas | CockroachDB / Vitess | Neki |
|---|---|---|---|---|
| Application transparency | No | Yes | Partial | Yes |
| Zero-downtime scaling | No | No | Yes | Yes |
| Postgres compatibility | Yes | Yes | Partial | Yes |
| Deployment effort | Months | Hours | Weeks | Hours |
| Maintenance burden | Very high | Medium (scaling limits) | High | Low |
| Monthly cost (10TB) | $8k+ | $12k+ | $4k | $3k |
You pay for storage and compute. No surprises. No vendor lock-in.
$0.12 per GB/month + $0.08 per CPU-hour
$5k/month for 5 shards, then $500 per additional shard
Teams building consumer or enterprise SaaS who have hit the Postgres scaling wall and need to move fast:
Register interest
This is not a purchase and there is no card field. It puts your address, this product, and whatever you write below in front of a person, and you get a written answer about what finishing it, or handing it over for you to run yourself, would actually take.